报告题目：Statistical Tools for Assessing Agreement among Measurement Methods
报告人：Dr. Chia-Cheng Chen(陈家正博士),
       Institute of Statistical Science, Academia Sinica
报告时间：2015年1月8日(周四)上午9:30-10:30
报告地点：数学楼二楼学术报告厅
 
 
摘要：Assessing reliability among multiple measurement methods plays an important role in evaluating whether different observers give similar results of the measured responses. Among the statistical tools, intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC) are two popular indices for assessing the closeness between readings measured from different observers for continuous variables. In this talk, the concept of agreement for ICC and CCC will be introduced. In addition, we will present our proposed CCC-type index for data with repeated measurements rather than replications where both observers and times are treated as random effects. This index is especially used when the observer/time effect is treated as random (i.e. observers/times are randomly selected from the observer/time population). We use the method of moments approach for estimation and a modified bootstrap approach for inference. Finally, an image data is used to illustrate our proposed methodology.